Discovery of a new faint dwarf galaxy associated with NGC 253
D.J. Sand, D. Crnojević, J. Strader, E. Toloba, J.D. Simon, N. Caldwell, P. Guhathakurta, B. McLeod, A. C. Seth
aa r X i v : . [ a s t r o - ph . GA ] J un ApJL, in prep.
Preprint typeset using L A TEX style emulateapj v. 8/13/10
DISCOVERY OF A NEW FAINT DWARF GALAXY ASSOCIATED WITH NGC 253 * D. J. Sand, D. Crnojevi´c, J. Strader, E. Toloba,
J.D. Simon, N. Caldwell, P. Guhathakurta, B. McLeod, A. C. Seth ApJL, in prep.
ABSTRACTWe report the discovery of a new faint dwarf galaxy, which we dub Scl-MM-Dw1, at a projecteddistance of ∼
65 kpc from the spiral galaxy NGC 253. The discovery results from the PanoramicImaging Survey of Centaurus and Sculptor (PISCeS), a program with the Magellan/Megacam imagerto study faint substructure in resolved stellar light around massive galaxies outside of the Local Group.We measure a tip of the red giant branch distance to Scl-MM-Dw1 of D =3.9 ± &
10 Gyr, [Fe/H] ∼− ∼
500 Myr. Scl-MM-Dw1 has a half-light radius of r h =340 ±
50 pc and an absolutemagnitude of M V = − ± Subject headings: later INTRODUCTION
The Λ+Cold Dark Matter (ΛCDM) model for struc-ture formation is successful on large scales ( &
10 Mpc; e.g.Efstathiou et al. 1992; Jaffe et al. 2001; Percival et al.2001; Spergel et al. 2007). In this canonical model,galaxies grow hierarchically within dark matter halos(e.g. Springel et al. 2006). However, a quantitative ver-ification of this model has met with extensive chal-lenges, particularly on galaxy scales and smaller. Forinstance, comparisons between the number of subha-los seen in numerical simulations and the number ofMilky Way or M31 dwarf galaxies show a factor of >
100 discrepancy (the “missing satellites problem”, e.g.Klypin et al. 1999; Moore et al. 1999). Further, kine-matic studies of the most luminous Local Group dwarfsindicate that their densities are lower than those of sim-ulated halos of the same mass (the “too big to fail” prob-lem; Boylan-Kolchin et al. 2011, 2012). These small-scale tests of ΛCDM galaxy formation models have beenmostly confined to substructure around the Milky Wayand M31.Two spiral galaxy halos in a loose group do not sam-ple the full spectrum of galaxy masses, morphologies andenvironments necessary to have a complete picture ofgalaxy assembly. In particular, we need to determine [email protected] * This paper includes data gathered with the 6.5 meter Mag-ellan Telescopes located at Las Campanas Observatory, Chile. Texas Tech University, Physics Department, Box 41051,Lubbock, TX 79409-1051, USA Michigan State University, Department of Physics and As-tronomy, East Lansing, MI 48824, USA UCO/Lick Observatory, University of California, SantaCruz, 1156 High Street, Santa Cruz, CA 95064, USA Observatories of the Carnegie Institution for Science, 813Santa Barbara Street, Pasadena, CA 91101, USA Harvard-Smithsonian Center for Astrophysics, Cambridge,MA 02138, USA Department of Physics and Astronomy, University of Utah,Salt Lake City, UT 84112, USA whether the Local Group just happens to be an out-lier in terms of its faint dwarf galaxy population, or ifsubstructure in other galaxies presents similar tensionwith ΛCDM. Pioneering studies of other galaxies havebegun, including resolved stellar population work in M81(Chiboucas et al. 2009, 2013) and low surface brightnesssearches out to larger distances (e.g. Merritt et al. 2014).In this work, we present the discovery of the faintdwarf galaxy Scl-MM-Dw1, the first result of a newsurvey to study faint substructure around NGC 253(a star-bursting spiral galaxy in the Sculptor group)and NGC 5128 (the dominant elliptical galaxy in theCentaurus A group) in order to broaden our observa-tional knowledge of faint substructure beyond the LocalGroup. In §
2, we briefly summarize the strategy forour panoramic imaging survey of nearby galaxies. Wepresent our data reduction methodology in § §
4. We discussand conclude in § SURVEY DESCRIPTION & DISCOVERY OFSCL-MM-DW1
We have begun the Panoramic Imaging Survey of Cen-taurus and Sculptor (PISCeS), a resolved stellar surveyof two of our nearest massive galaxy neighbors —NGC253 and NGC 5128—to search for archaeological rem-nants of their buildup via tidal streams and satellitedwarf galaxies down to M V ∼ − M V ∼−
14 mag (Cote et al. 1997;Jerjen et al. 2000; Karachentsev et al. 2002, 2004). Ourgoal is to survey each galaxy out to r =150 kpc to allowa direct comparison with the PAndAS survey in M31(McConnachie et al. 2009), and recent work around theMilky Way.To accomplish this ambitious task, we are utilizing theMegacam instrument (McLeod et al. 2006), which hasa total field of view of ∼ ×
24’ at the f /5 focus on Sand et al.the Magellan Clay telescope. The outer survey radiusof ∼
150 kpc amounts to ∼
21 square degrees on the skyfor NGC 253/NGC 5128 (both at D ∼ ∼
130 Megacam pointings for each halo. In roughaccordance with previous dwarf galaxy discoveries in theSculptor group, we will label our new dwarfs numericallystarting with Scl-MM-Dw1, where the “MM” denotes theMagellan/Megacam instrumentation utilized by our pro-gram.Imaging is done in two bands, g and r , chosen forsurvey speed and so that new dwarf candidates can beclearly identified and initial properties (e.g. distance,size, approximate star formation history, and luminos-ity) ascertained from their resolved stellar populationsduring the course of the survey. At the distance toNGC 253/NGC 5128, the tip of the red giant branch(TRGB) is at r ∼ M V ∼− r , g ∼ . . ′′ . ′′ . ′′
6; Figure 1). This new dwarf galaxy, which wedub Scl-MM-Dw1, is ∼
65 kpc away from the center ofNGC 253 in projection. Scl-MM-Dw1 is just visible inDigital Sky Survey archival images, but at a level wherea detection is not secure, especially due to the presenceof a handful of bright foreground stars and backgroundgalaxies in its vicinity. DATA REDUCTION
The data presented in this paper were taken on 2013September 2 and 7 (UT) with Magellan/Megacam. Pho-tometric conditions were variable, but the seeing was ex-cellent. The final stacked images consisted of 15 × g and r band, with image pointspread functions (PSFs) of 0 . ′′ . ′′
6. Initial data re-duction, including standard image detrending, astromet-ric matching and stacking is performed by the Smith-sonian Astrophysical Observatory Telescope Data Cen-ter utilizing a code developed by M. Conroy, J. Rolland B. McLeod. Stellar photometry on the final im-age stacks was performed using a methodology similarto that of previous work utilizing Magellan/Megacamfor resolved stellar studies (Sand et al. 2012) with the
DAOPHOTII/Allstar package (Stetson 1994).Our Fall 2013 data of Scl-MM-Dw1 were not taken inphotometric conditions, and so we utilized data takenin Fall 2012 of the same field in clear conditions (butpoor seeing) to calibrate the presented data. We con-vert instrumental magnitudes into the SDSS photometricsystem by observing fields covered by SDSS at differentairmasses during clear nights, which allowed for an at-mospheric extinction term to be measured. Zeropointsand color terms were similar to our previous work withMagellan/Megacam (Sand et al. 2012).Once the Fall 2013 data was calibrated, we performeda series of artificial star tests (utilizing the
DAOPHOT routine
ADDSTAR ) to calculate our photometric errorsand completeness as a function of magnitude and colorfor the field. Since crowding is not an issue, artificial stars were placed into our images on a regular grid whosespacing was several times that of the stellar PSF. Overten iterations, we injected ∼ × artificial stars intoour data, covering a r -band magnitude range of 18 to30 and a g − r color of − r =25.6 (24.4) and g =26.5 (25.3). These limits areslightly shallower than our ultimate goals due to the skyconditions; data taken on clear, good-seeing nights dohit our magnitude targets as described in §
2. Thereis no indication that the unresolved light from Scl-MM-Dw1 changes our completeness or uncertainties in thatregion.Our final catalog was corrected for Galactic extinc-tion (Schlafly & Finkbeiner 2011), and all magnitudesreported in the remainder of this paper will be correctedin this way. Figure 1 shows our color magnitude diagram(CMD) of Scl-MM-Dw1, where the error bars show thetypical magnitude and color uncertainty at different r -band magnitude levels, and a g − r of 0.8 mag. PROPERTIES OF SCL-MM-DW1
In this section, we determine the basic properties ofScl-MM-Dw1. We lead the discussion with its stellarpopulations, as these are critical for the subsections thatfollow: the distance, structure, and luminosity of thenewly discovered dwarf.
Stellar Population
Close inspection of the CMD as seen in Figure 1 hintsat two interpretations for the stellar population and ap-proximate distance to Scl-MM-Dw1, and we critically as-sess each here.The first scenario—which we ultimately argueagainst—places Scl-MM-Dw1 nearer than the nominaldistance to NGC 253 (with a difference in distance mod-ulus of ∆ µ ∼ µ ∼ D =2.6 Mpc), asshown in Figure 2. In the left panel of Figure 2 weover-plot a 10 Gyr, [Fe/H]= − r ≈ g − r ≈ r ∼
24 mag is statisticallysignificant. An RGB luminosity function would not havesuch a gap, which suggests that the true dwarf distanceis further away, and that the secondary peak is not theTRGB, but a different stellar population associated withthe dwarf.This leads to the second scenario, where Scl-MM-Dw1is consistent with being at the distance of NGC 253(we directly measure the distance to Scl-MM-Dw1 viaits TRGB in § − . − ∼
500 Myr.We are confident that at least some of the stars in thissecondary peak in the CMD/luminosity function are as-sociated with Scl-MM-Dw1 because they appear to bespatially clustered as the RGB stars are. In the rightpanel of Figure 3, we show a spatial map of stars consis-tent with the RGB and AGB stellar populations in thisscenario. Both populations show a clear overdensity atthe position of the dwarf, reaffirming their significance.Only a broad view of the stellar population in Scl-MM-Dw1 can be gleaned from the current data set. Over-all, the CMD of Scl-MM-Dw1 is consistent with an old( &
10 Gyr), metal poor ([Fe/H] ∼−
2) stellar populationat roughly the distance to NGC 253 (see § ∼
500 Myr) stellar populationas well. A low level of even more recent star forma-tion could also be present, but would not be detectedgiven our current image depths. A similar CMD mor-phology (with an RGB, ‘gap’ and second brighter stel-lar population) is often seen in the literature for otherdwarfs. Indeed, inspection of the CMD compilation ofDalcanton et al. (2009) shows several dwarfs with a sim-ilar morphology (including DDO44/KK61, KDG2/E540-30, and KDG61/KK81) where an old RGB stellar pop-ulation exists with a younger AGB stellar population.Deeper optical and near-infrared imaging will be able tofurther constrain Scl-MM-Dw1’s stellar population.
The Distance to Scl-MM-Dw1
We measure the distance of Scl-MM-Dw1 with theTRGB method (e.g., Lee et al. 1993; Rizzi et al. 2007).For metal-poor populations, the TRGB has a constant I -band absolute magnitude, making it a widely used dis-tance indicator. However, to our best knowledge thereis no empirical TRGB calibration for SDSS bands (e.g.,Bellazzini 2008). We thus adopt the Dartmouth stel-lar evolutionary models (Dotter et al. 2008) and derivethe theoretical absolute TRGB magnitude in r -band us-ing isochrones with a fixed age of 12 Gyr and metal-licities ranging from [Fe/H]= − . − .
0, obtaining M T RGBr = − . ± .
1. We apply a Sobel edge-detectionfilter (Lee et al. 1993) to the luminosity function of Scl-MM-Dw1 with a color cut of 0 . < ( g − r ) < . r T RGB = 24 . ± .
26, which translates into a distancemodulus of m − M =27 . ± .
32 with our TRGB cal-ibration. The uncertainties reflect the photometric er-rors and the sparseness of the luminosity function. As atest, we also compute r T RGB for all stars in the Mega-cam pointing not belonging to Scl-MM-Dw1, i.e., prob-able NGC 253 halo stars. We find 24 . ± .
24 whichgives a distance modulus of m − M =27 . ± .
30. Theexcellent agreement with literature estimates for the dis-tance of NGC 253 (e.g., Radburn-Smith et al. 2011, whofound m − M =27.70 ± Structure
Despite its distance and relatively small number of re-solved stars, we can determine many of the structuralparameters of Scl-MM-Dw1 through standard meth-ods. We use a maximum likelihood technique for con-straining structural parameters, based on the recipeof Martin et al. (2008), and identical to that done inSand et al. (2012). We cut out a 10 ×
10 arcmin regionaround the approximate center of the dwarf for use inour calculations. The stars selected for the structuralanalysis are those in the boxed regions of Figure 3 (leftpanel), corresponding to the RGB and AGB star can-didates in Scl-MM-Dw1. We fit a standard exponentialprofile plus constant background to the data, with thefollowing free parameters: the central position ( α , δ ),position angle (PA; θ ), ellipticity ( ǫ ), half-light radius( r h ) and background surface density (Σ b ). Uncertaintieson structural parameters were determined via bootstrapresamples, from which 68% confidence limits were calcu-lated.Our results are presented in Table 1. The half lightradius of Scl-MM-Dw1 is 340 ±
50 pc, which is compara-ble to the size of the MW’s classical dwarf spheroidals.The central position of Scl-MM-Dw1 is also well con-strained. The ellipticity (and thus position angle) ofScl-MM-Dw1 is not well constrained, and our maximumlikelihood analysis can provide only an upper limit of ǫ< ∼ Luminosity
We estimate Scl-MM-Dw1’s luminosity directly fromaperture photometry. We use an aperture radius equalto the half-light radius (16 . ′′ M g = − ± M r = − ± § M r , M g magnitudes to M V , we use thefilter transformation equations of Jordi et al. (2006), andfind M V = − ± DISCUSSION AND CONCLUSIONS
We have presented the discovery of Scl-MM-Dw1, afaint dwarf galaxy found ∼
65 kpc in projection fromNGC 253. Its inferred distance (3.9 ± >
10 Gyr; [Fe/H] ∼−
2) and AGB stars with Sand et al.an age of ∼
500 Myr. Dwarfs with recent star forma-tion are rarely seen within ∼
250 kpc of massive par-ent galaxies such as NGC 253 (e.g. Grebel et al. 2003;Grcevich & Putman 2009, among many others), so animproved distance to the dwarf would be of significantinterest. Scl-MM-Dw1 is not detected in the HI ParkesAll-Sky Survey (HIPASS; Barnes et al. 2001), with a 3- σ HI gas mass upper limit of M HI . × M ⊙ .The size and luminosity of Scl-MM-Dw1 are compara-ble to those of the MW dwarf galaxies, as can be seenin Figure 4. The MW satellites most similar to Scl-MM-Dw1 are Sculptor ( M V = − r h =283 pc) and Ca-rina ( M V = − r h =250 pc). Carina in particularalso has evidence for young stellar populations ( ∼ r .
150 kpc of bothNGC 253 and NGC 5128. Based on initial analyses, ourdataset will be able to detect all dwarf galaxies down to M V ∼− REFERENCESBarnes, D. G., Staveley-Smith, L., de Blok, W. J. G., et al. 2001,MNRAS, 322, 486Bellazzini, M. 2008, Mem. Soc. Astron. Italiana, 79, 440Boylan-Kolchin, M., Bullock, J. S., & Kaplinghat, M. 2011,MNRAS, 415, L40—. 2012, MNRAS, 422, 1203Chiboucas, K., Jacobs, B. A., Tully, R. B., & Karachentsev, I. D.2013, AJ, 146, 126Chiboucas, K., Karachentsev, I. D., & Tully, R. B. 2009, AJ, 137,3009Cote, S., Freeman, K. C., Carignan, C., & Quinn, P. J. 1997, AJ,114, 1313Dalcanton, J. J., Williams, B. F., Seth, A. C., et al. 2009, ApJS,183, 67Dotter, A., Chaboyer, B., Jevremovi´c, D., et al. 2008, ApJS, 178,89Efstathiou, G., Bond, J. R., & White, S. D. M. 1992, MNRAS,258, 1PGirardi, L., Williams, B. F., Gilbert, K. M., et al. 2010, ApJ, 724,1030Grcevich, J., & Putman, M. E. 2009, ApJ, 696, 385Grebel, E. K., Gallagher, III, J. S., & Harbeck, D. 2003, AJ, 125,1926Jaffe, A. H., Ade, P. A., Balbi, A., et al. 2001, Physical ReviewLetters, 86, 3475Jerjen, H., Binggeli, B., & Freeman, K. C. 2000, AJ, 119, 593Jordi, K., Grebel, E. K., & Ammon, K. 2006, A&A, 460, 339Karachentsev, I. D., Karachentseva, V. E., Huchtmeier, W. K., &Makarov, D. I. 2004, AJ, 127, 2031 Karachentsev, I. D., Sharina, M. E., Dolphin, A. E., et al. 2002,A&A, 385, 21Klypin, A., Kravtsov, A. V., Valenzuela, O., & Prada, F. 1999,ApJ, 522, 82Koposov, S., Belokurov, V., Evans, N. W., et al. 2008, ApJ, 686,279Lee, M. G., Freedman, W. L., & Madore, B. F. 1993, ApJ, 417,553Martin, N. F., de Jong, J. T. A., & Rix, H.-W. 2008, ApJ, 684,1075McConnachie, A. W. 2012, AJ, 144, 4McConnachie, A. W., Irwin, M. J., Ibata, R. A., et al. 2009,Nature, 461, 66McLeod, B., Geary, J., Ordway, M., et al. 2006, in ScientificDetectors for Astronomy 2005, ed. J. E. Beletic, J. W. Beletic,& P. Amico, 337Merritt, A., van Dokkum, P., & Abraham, R. 2014, ApJ, 787, L37Monelli, M., Pulone, L., Corsi, C. E., et al. 2003, AJ, 126, 218Moore, B., Ghigna, S., Governato, F., et al. 1999, ApJ, 524, L19Percival, W. J., Baugh, C. M., Bland-Hawthorn, J., et al. 2001,MNRAS, 327, 1297Radburn-Smith, D. J., de Jong, R. S., Seth, A. C., et al. 2011,ApJS, 195, 18Rizzi, L., Tully, R. B., Makarov, D., et al. 2007, ApJ, 661, 815Sand, D. J., Strader, J., Willman, B., et al. 2012, ApJ, 756, 79Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103Spergel, D. N., Bean, R., Dor´e, O., et al. 2007, ApJS, 170, 377Springel, V., Frenk, C. S., & White, S. D. M. 2006, Nature, 440,1137Stetson, P. B. 1994, PASP, 106, 250Walsh, S. M., Willman, B., & Jerjen, H. 2009, AJ, 137, 450 new NGC 253 dwarf galaxy 5
Fig. 1.—
Top:
Scl-MM-Dw1 as seen in our g -band image. The top left panel shows a zoomed out view, with the black arrow indicatingthe direction towards NGC 253 (nearly due North) and ∼
65 kpc away in projection. There is also a clear surface brightness enhancementassociated with the dwarf. The top right panel shows a zoomed in view, with many resolved stars apparent. Note also the spiral/S0 galaxyand a bright foreground star near the center of Scl-MM-Dw1. The radius of the circle is 0.45 arcmin, which is the region from which theCMD in the bottom panel is drawn.
Bottom:
The bottom left panel shows the CMD of Scl-MM-Dw1 within the 0.45 arcminute radiuscircle drawn. There is a clear overdensity of stars in this region, with a morphology similar to a RGB, which we analyze in more detailin § r -band magnitudes, as determined via our artificial startests. The blue dashed line shows the 50% completeness limit. The three adjacent panels show CMDs from random equal-area regions ofthe same Megacam field, illustrating typical “background” CMDs. Sand et al.
Fig. 2.—
The “nearby” interpretation of Scl-MM-Dw1, which we ultimately reject.
Left:
A CMD of Scl-MM-Dw1, identical to thatshown in Fig 1, is shown. The blue theoretical isochrone is of a 10 Gyr, [Fe/H]= − µ =27.1 ( D =2.6 Mpc), ∆ µ ∼ r ≈ g − r ≈ Right:
Luminosity function of stars consistent with being a part of the RGB of Scl-MM-Dw1, if it was at a distance of µ =27.1.The dotted points show the luminosity function of equal-area random pointings throughout the field with the error bars spanning 95% ofthe random draws (see § r ∼
24 is statistically significant; thestars associated with this secondary peak are also clustered around the position of Scl-MM-Dw1, see Figure 3. The number counts of starsin the gap between the two peaks in the luminosity function are consistent with being background. Based on this, we are confident thatScl-MM-Dw1 is not a dwarf galaxy at µ =27.1 ( D =2.6 Mpc), but that the secondary peak is of an AGB stellar population. new NGC 253 dwarf galaxy 7 Fig. 3.—
The NGC 253 satellite interpretation of Scl-MM-Dw1.
Left:
The CMD as plotted in Figures 1 & 2. The orange and purpleisochrones are for a 400 and 630 Myr stellar population with a [Fe/H]= −
1. The blue isochrone is of a 13 Gyr, [Fe/H]= − Right:
Spatial map of stars consistent with being AGB (red; red selection box in the left panel) and RGB (black; blackselection box in the left panel) stars. There is an overdensity of stars of both populations at the position of the dwarf.
TABLE 1Properties of Scl-MM-Dw1
Parameter Value m − M (mag) 27.98 ± ± M V (mag) − ± ± ± r h (arcsec) 16.8 ± r h (pc) 340 ± ǫ < aa ǫ corresponds to the 95% upper confidence limit. Sand et al.